API tools
Choose an operation and body, then generate a validated JSON sample.
Your specification and payloads stay in this browser; no API request is sent.
Reading operations…
Use the OpenAPI mock generator when frontend tests need the shape of an API response before the endpoint is ready. Keep the contract beside the fixture so a status-code change does not quietly leave a success payload in an error test.
Generate JSON examples from OpenAPI without installing a mock server. The selected operation determines the body schema; quantity controls independent payloads, not the number of elements inside a response array.
The OpenAPI mock generator lets you choose a declared 400 response just as directly as 200. Switch between named examples to test different documented outcomes without editing the definition or remembering which payload belongs to which status.
Turn off example priority to explore the supported schema with a fixed seed. Identical input and options produce the same batch for a given engine version; a seed change never invents extra variety in an explicit constant or example.
The initial YAML example is parsed into selectors automatically. Choose request to generate requestBody only; response uses the selected status and media type. Parameters in the URL, query or headers are outside this body-only workflow.
Named media examples have priority over schema and property examples. Disable example priority for synthetic data. An invalid example stops generation instead of silently falling back, so fix it or deliberately switch to schema sampling.
One JSON sample keeps its original body shape. A batch adds an outer array; JSONL puts one complete body on each line. Ctrl/Cmd + Enter generates manually, and changing a selector disables exports until the next successful run.
Inspect each method and path, select a request or response, and choose any declared numeric response, range or default. The initial choice favors 200, another 2xx response, then the first available entry. Mock API responses need not all be success cases.
Media example values, named examples and same-document Example Object references are supported. Root and nested schema examples are used when enabled and applicable. Every exported payload is checked against the supported schema after request/response field filtering.
Request samples exclude readOnly properties; response samples exclude writeOnly properties. Their required entries are removed for that direction. Supplied examples containing an excluded property fail rather than being rewritten behind your back.
Create 1–100 independent payloads for OpenAPI test data. The OpenAPI mock data generator shares its bounded sampler with the schema tool. Seed and count are explicit; uniqueItems applies within each payload array, not across the whole batch.
Paste the complete supported document and select the operation, direction and media type. Keep example priority enabled for documented payloads, or disable it to sample the schema. A standalone schema belongs in the dedicated schema data generator.
Select the response status and JSON media type first, then choose its named example. Inline example values and local Example Object references are accepted. Media examples take precedence over schema examples; examples from another status are never mixed in.
Yes, when the operation declares a matching response with a supported JSON schema. Select 400, 404 or another declared status. This creates that body as a fixture; it does not send an HTTP response or simulate a server failure.
The default seed is 42. Keeping the document and options fixed reproduces schema sampling, while explicit examples, const and single-value enums can remain unchanged for every seed. Disable example priority before expecting seed-dependent variation.
No. Validation checks the supplied or generated body against the supported effective schema. It does not contact an API, execute middleware or verify HTTP headers, authentication and business rules.
See a constraint we should support?
Start with one payload in the OpenAPI mock generator, inspect its source and status selection, then increase the count only when the test needs a batch.